Search in the Abstract Database

Search Abstracts 2015

* = Presenting author

P284 Surveillance for dysplasia in IBD: are we doing it right? Results from our daily practice

D. Trabulo*1, J. Moleiro2, I. Rosa2, C. Fidalgo3, J. Pereira da Silva2, P. Fidalgo4, R. Fonseca5, P. Chaves5, A. Dias Pereira2

1Centro Hospitalar de Setúbal - Hospital de São Bernardo, Gastroenterology, Setúbal, Portugal, 2Instituto Português de Oncologia de Lisboa Francisco Gentil, EPE, Gastroenterology, Lisboa, Portugal, 3Hospital Beatriz Ângelo, Gastroenterology, Loures, Portugal, 4Fundacao Champallimaud, Gastroenterology, Lisboa, Portugal, 5Instituto Português de Oncologia de Lisboa Francisco Gentil, EPE, Pathology, Lisboa, Portugal


Dysplasia is a known complication of long-standing Inflammatory Bowel Disease (IBD). Guidelines for surveillance were recently updated, but their applicability in daily clinical practice is largely unknown. We planned to survey our regular clinical practice in the light of these new guidelines.


From an IBD clinic with 202 registered patients, we evaluated 59 patients under dysplasia surveillance, in a 3 year period. Demographic, clinical, endoscopic and histological data were collected from patient files.


In 59 patients, 66 colonoscopies under sedation were performed during the studied period: 50% were females; 54 years-old mean-age at study entry; 88% had ulcerative colitis (79% pancolitis) and 12% had Crohn's colitis (CC). Mean duration of disease was 16 years (8-32). One patient had primary sclerosing cholangitis and 8 had family history of colorectal cancer; 74% of patients had one risk factor for dysplasia in IBD and 13% had 2 or more; 89% were using aminosalicilates, 32% thiopurines and 12% anti-TNF therapy. Nine UC patients had endoscopic activity (Mayo subscore ≥ 2) and 2 others had only histological active inflammation. One CC patient had active colonic disease, both by endoscopy and histology. Macroscopic lesions were found in 61% of patients by conventional colonoscopy. 44% patients had adenoma-like lesions (mean size 5 mm) - from these, only 45% had low-grade dysplasia at histology. Two patients had non-adenoma-like lesions (10 and 15 mm): one was a tubulovillous adenoma with low-grade dysplasia and the other was a hyperplasic polyp. The mean number of random biopsies was 29 (16-40). Low-grade dysplasia at random biopsies was found in 1 patient. No high-grade dysplasia or adenocarcinomas were found. Chromoendoscopy was only possible in 11 patients, due to inadequate bowel cleansing in all others. Lesions were found in 7 patients; in 2 of them, lesions with low-grade dysplasia had been missed by conventional colonoscopy.


The majority of patients had at least one risk factor justifying colonoscopic surveillance. Most of them were receiving aminosalicilates as recommended. However, the mean number of random biopsies was below standards. Moreover, chromoendoscopy was only performed in a minority of patients, but it identified 2 cases of dysplasia missed by conventional colonoscopy. In conclusion, IBD dysplasia surveillance in community practice has practical limitations, but chromoendoscopy must be endorsed, as well as patient education for optimized bowel preparation.